[SOLVED] Issues with syncing

maenda

Active Member
Jul 10, 2020
36
0
26
50
Hi all,

The status of one of the nodes keeps pending on SYNCING and is not going forward.
Checking the logs, gives me the following:

Code:
Jul 23 09:55:51 mx-2 pmgmirror[675]: starting cluster syncronization
Jul 23 09:55:51 mx-2 pmgmirror[675]: syncing deleted node 2 from master 'IP'
Jul 23 09:55:51 mx-2 pmgmirror[675]: database sync 'mx-1' failed - DBD::Pg::st execute failed: ERROR:  duplicate key value violates unique constraint "cmailstore_pkey"#012DETAIL:  Key (cid, rid)=(2, 8) already exists. at /usr/share/perl5/PMG/DBTools.pm line 1086.
Jul 23 09:55:52 mx-2 pmgmirror[675]: cluster syncronization finished  (1 errors, 1.60 seconds (files 0.14, database 1.30, config 0.16))


Any idea how to fix this?

Thanks
 
Last edited:
What were the steps you did to get there? - I assume you have a clustered environment and deleted nodes?!
what's the output of:
`pmgcm status`
 
Hi Stoiko,

The node has been replaced indeed because of the sync problems. The master was not touched. I removed the node by the PMGM DELETE id command.
Also, a letsencrypt certificate has been installed. But I guess this should be no problem?

Output:

Code:
pmgcm status
NAME(CID)--------------IPADDRESS----ROLE-STATE---------UPTIME---LOAD----MEM---DISK
mx-1(1)              ip_of_master  master A           22:57   0.18    31%     7%
mx-2(6)              ip_of_node    node   S           22:55   0.18    26%    11%
root@mx-1:~#
 
I removed the node by the PMGM DELETE id command
did you setup the new node freshly, before adding it to the cluster?
if not - try that.
(keep a backup of the current node and one of the master just in case)
 
delete it - check that it is indeed gone from the cluster configuration (/etc/pmg/cluster.conf) then use a fresh install and add that to the cluster